Analyzing Profile and Data

Now that we have all these profiles, what should we do with them? Can users take advantage of profiles or are they only tools for webmasters and marketing professionals? Which tools can be used to mine the profiles?

Commerce Server Analysis Tools

Commerce Server 2002 offers the richest set of profile analytical tools in the Microsoft portal platform, as befits this flagship e-business product. The Commerce Server Business Desk is the user interface framework that supports analysis of profiles and other site data. It is a web interface that is separate from the commerce site and exists for the benefit of the "back office" rather than shoppers. The Business Desk consists largely of Active Server Pages along with shared modules. It is designed to be extended by developers and customized for the needs of a particular business. A Harley-Davidson dealership handles its customers and products in a different way from a haberdasher or a food emporium, for instance. Therefore, each instance of Commerce Server has it own Business Desk.

The Business Desk is organized into list pages and edit pages. Because Commerce Server is largely parameter-driven by data from SQL Server, Business Desk provides a user interface to access both global and module-specific settings. For instance, user profiles can be accessed through the Business Desk. You can search for users in the Users module based on fields in the profile such as logon name, email address, first name, last name , and custom properties.

NOTE The profiles are a rich source of information on the users of the portal, as well as the basis of analysis and decisions relating to content, marketing, and outreach. Before you unleash this data for all to see, however, you need to be aware of the confidentiality of some profile information. Be sure to follow advice of your legal counsel and to adhere to your site's privacy guidelines.

Commerce Server 2002 comes with 35 preconfigured reports covering such areas as sales, web usage, advertising, and user reports. User reports include distinct users by time, new registered users, user days to register, and user trends. You can customize these reports or add new reports of your own design. The reports are run from a data warehouse that is generated from site data, including web logs. The data is stored in online analytical processing (OLAP) cubes, a format that dramatically improves performance for aggregations. You can also build your own custom reports against the OLAP cubes from your Commerce site, with tools such as Excel pivot tables.

You can also create your own custom reports using Commerce Server data. These reports come from the OLAP cube created by Commerce Server. Table 7.6 lists the Commerce Server Data Warehouse cubes and associated dimensions, along with the source tables that contain that data.

Table 7.6. Commerce Server Cubes and Dimensions

Cube	Dimensions	Source tables	Reports
Bandwidth	DateHour Day of Week Hour of Day Site	HitsByHour	Bandwidth Summary Bandwidth Trends
Basket Events	DateHour Events ProductCategory Site UserType	csdw_BasketEventsFactView	Shopping Basket Events
Buyer Visits	BuyerVisits Site UserType	UserBrowsingtoPurchaseFactView	Buyer Browse to Purchase
Entry Pages	Date Site URI	FirstUriByDate	Entry Pages
Exit Pages	Date Site URI	LastUriByDate	Exit Pages
mscsCampaign	Advertiser DateHour Events Page Group Site UserType	mscsCampaignEvents mscsOrderedImpressions	Ad Reach and Frequency by Date Ad Reach and Frequency per Advertiser Campaign Event Summary Campaign Item Summary
MscsCampaignEvents	Advertiser DateHour Events Page Group Site User Type	dbo.csdw_CampaignEventsFactView
MscsDateIntersect	Date Hour of Day Day of Week Week of Year	csdw_DateFactView
mscsOrderedImpressions	Advertiser Site	dbo.csdw_OrderedImpressionsFactView
mscsSales	Address Currency Date ProductCategory Site UserType	dbo.csdw_SalesFactView	Product Sales
mscsVisitInfo	BrowserOS BrowserType Date DateHour Day of Week Hour of Day Referrer Site UserType Week of Year	dbo.csdw_VisitInfoFactView
mscsWebUsage	BrowserOS BrowserType Date DateHour Day of Week Hour of Day Referrer Site UserType Week of Year	mscsDateIntersect mscsVisitInfo	Activity by Browser Distinct Users By Time General Activity Statistics Top Referring Domains by Request Usage Summary by Day of Week Usage Summary by Hour of Day Usage Summary by Week of Year Usage Trends User Trends User Visit Trends
OrderEvents	DateHour Day of Week Events Hour of Day Site UserType Week of Year	Order	Order Events
Page Usage	Date HTTP Status Is Request Level Directory Site URI Win32Status	HitsInfo	Directories Hits by HTTP Status Hits by Win32 Status Top Requested Pages

SharePoint and SQL Server Profile Analysis

If you choose to develop a custom profile in SQL Server, you will not have the benefit of the canned reports or the off-the-shelf data warehouse capabilities of Commerce Server, but you can still write your own reports against the SQL Server database. This is also true for SharePoint Portal Server. Because it stores its profile data in SQL Server, you can use your favorite SQL Server reporting and data-mining tools to extract this data.